[
https://issues.apache.org/jira/browse/HIVE-12737?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Gopal V updated HIVE-12737:
---------------------------
Description:
{code}
private void readHeader() throws IOException {
....
ByteBuffer slice = this.slice(chunkLength);
if (isOriginal) {
uncompressed = slice;
isUncompressedOriginal = true;
} else {
if (isUncompressedOriginal) {
uncompressed = allocateBuffer(bufferSize);
isUncompressedOriginal = false;
} else if (uncompressed == null) {
uncompressed = allocateBuffer(bufferSize);
} else {
uncompressed.clear();
}
codec.decompress(slice, uncompressed);
}
} else {
throw new IllegalStateException("Can't read header at " + this);
}
}
{code}
when reading the 2nd chunk, uncompressed didn't check the data left in itself.
chunkSize is set by hive.exec.orc.default.buffer.size, if user set a number
that can't mod 8, when read double, error occurs.
was:
private void readHeader() throws IOException {
....
ByteBuffer slice = this.slice(chunkLength);
if (isOriginal) {
uncompressed = slice;
isUncompressedOriginal = true;
} else {
if (isUncompressedOriginal) {
uncompressed = allocateBuffer(bufferSize);
isUncompressedOriginal = false;
} else if (uncompressed == null) {
uncompressed = allocateBuffer(bufferSize);
} else {
uncompressed.clear();
}
codec.decompress(slice, uncompressed);
}
} else {
throw new IllegalStateException("Can't read header at " + this);
}
}
when reading the 2nd chunk, uncompressed didn't check the data left in itself.
chunkSize is set by hive.exec.orc.default.buffer.size, if user set a number
that can't mod 8, when read double, error occurs.
> orc Instream.java CompressedStream.readHeader() can meet error when read
> double.
> --------------------------------------------------------------------------------
>
> Key: HIVE-12737
> URL: https://issues.apache.org/jira/browse/HIVE-12737
> Project: Hive
> Issue Type: Bug
> Components: ORC
> Affects Versions: 1.0.1
> Reporter: xuanflyer
> Priority: Minor
>
> {code}
> private void readHeader() throws IOException {
> ....
> ByteBuffer slice = this.slice(chunkLength);
> if (isOriginal) {
> uncompressed = slice;
> isUncompressedOriginal = true;
> } else {
> if (isUncompressedOriginal) {
> uncompressed = allocateBuffer(bufferSize);
> isUncompressedOriginal = false;
> } else if (uncompressed == null) {
> uncompressed = allocateBuffer(bufferSize);
> } else {
> uncompressed.clear();
> }
> codec.decompress(slice, uncompressed);
> }
> } else {
> throw new IllegalStateException("Can't read header at " + this);
> }
> }
> {code}
> when reading the 2nd chunk, uncompressed didn't check the data left in itself.
> chunkSize is set by hive.exec.orc.default.buffer.size, if user set a number
> that can't mod 8, when read double, error occurs.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
